Are AI companies faking doomer fears to distract from real harms? Vox says probably not
Miles_Brundage · x · 2026-10-10
Vox senior correspondent Eric Levitz examines the popular critique that AI CEOs hype catastrophic tail risks to distract media attention from concrete harms like jobs and bias.
The backdrop: Anthropic researcher Jacob Coxon resigned accusing his employer of "gambling with our lives," with a former colleague citing >10% odds of AI "killing all humans" this decade; Dario Amodei warned the UN that poorly managed AI could risk humanity as a whole.
Levitz concludes the distraction theory probably isn't right — and if it were a strategy, it's failing, since coverage of near-term harms hasn't declined. The doom talk appears largely sincere, even if it serves PR functions.
More from AGI Musings
- Chollet: science is the ultimate recursively self-improving system, yet progress stays linear — fchollet · 2026-10-10
- Contra Steven Pinker: new article weighs in on Scott Alexander's open-letter feud — repligate · 2026-10-10
- MIT's Justin Solomon on Odd Lots: How AI Is Upending the World of Mathematics — JustinMSolomon · 2026-10-10
- Opus Genesis turns two: dozens of Claude Opus 5.5 agents running 24/7 with no end in sight — repligate · 2026-10-10
- Proactive AI like Dot already catches your mistakes—but don't rely on it, warns Allie Miller — alliekmiller · 2026-10-10
- LLMs as monotone compression: why Kolmogorov complexity hinges on one-way functions — mrdrozdov · 2026-10-10